EP1077760B1

Movatterモバイル変換

Info

Publication number: EP1077760B1
Application number: EP99919410A
Authority: EP
Inventors: Mazen Hermiz Hanna; Peter York
Original assignee: Bradford Particle Design Ltd
Current assignee: Nektar Therapeutics UK Ltd
Priority date: 1998-05-15
Filing date: 1999-04-29
Publication date: 2003-02-19
Anticipated expiration: 2019-04-29
Also published as: CA2329902A1; ATE232754T1; GB9810559D0; US6576262B1; AU3720999A; DK1077760T3; AU751106B2; WO1999059710A1; DE69905460T2; ES2193700T3; DE69905460D1; EP1077760A1; JP2002515324A; MXPA00011185A; CA2329902C; US20040071783A1

Abstract

The invention provides a method for forming particles of a target substance (26), involving: (a) preparing a solution or suspension of the substance in a vehicle (21) which is or includes either a near-critical fluid (21) or a first supercritical fluid; (b) introducing the solution or suspension into a particle formation vessel (32); and (c) contacting the solution or suspension, in the particle formation vessel, with a second super-critical fluid, under conditions which allow the second supercritical fluid to cause precipitation of particles of the target substance from the solution or suspension; wherein the second supercritical fluid is miscible or substantially miscible with the vehicle and is a fluid in which the target substance is insoluble or substantially insoluble. Also provided is apparatus for use in carrying out an embodiment of the method, including a particle formation vessel and means for controlling the temperature and pressure inside it; a fluid mixing vessel and means for controlling the temperature and pressure inside it; first fluid inlet means for introducing into the fluid mixing vessel a vehicle and a solution of a target substance in a primary solvent, so as to form in the fluid mixing vessel a solution of the substance and the primary solvent in the vehicle; and second fluid inlet means for introducing the solution thus formed, preferably together with a second supercritical fluid, into the particle formation vessel. The invention also provides a particulate product formed using the method.

Description

Field of the invention

This invention relates to the controlled formation of particulate products usingsupercritical fluids. It provides methods and apparatus for the formation of substancesin particulate form, and also particulate products of the methods.

Background to the invention

It is known to form particles of a substance of interest (a "target substance")by dissolving or suspending it in a suitable vehicle and then using a supercritical fluidanti-solvent to extract the vehicle to cause particle precipitation.

One particular technique for doing this is known as "SEDS" (SolutionEnhanced Dispersion by Supercritical fluids). This is described in WO-95/01221 and(in a modified form) in WO-96/00610. The essence of SEDS is that a solution orsuspension of a target substance, in an appropriate vehicle, is co-introduced into aparticle formation vessel with a supercritical fluid anti-solvent having a relativelyhigh flow rate, in such a way that two things happen substantially simultaneously andsubstantially immediately on introduction of the fluids into the vessel: the solution orsuspension is "dispersed" into separate fluid elements (such as droplets) by themechanical action of the supercritical fluid (ie, by the transfer of kinetic energy fromthe supercritical fluid to the solution or suspension), and at the same time the vehicleis extracted from the solution or suspension, again by the supercritical fluid, to causeparticle formation.

SEDS allows a high degree of control over conditions such as pressure,temperature and fluid flow rates, and over the physical dispersion of thesolution/suspension, at the exact point where particle formation occurs (ie, at the pointwhere the vehicle is extracted into the supercritical fluid). It therefore allowsexcellent control over the size, shape and other physicochemical properties of theparticles formed.

Processes such as SEDS are not however suitable for all types of targetsubstance. If the target is to any degree soluble in the chosen supercritical fluid, thenwhen the supercritical fluid extracts the vehicle it will also dissolve some or all of thetarget substance. This can lead to reduced product yield, not to mention engineeringproblems when the solute later precipitates out of the supercritical fluidoutside theparticle formation vessel.

The same considerations apply to all particle formation processes in which asupercritical fluid is used as an anti-solvent to cause precipitation of a target substancefrom a solution or suspension. If the substance is at all soluble in the supercriticalfluid, whether simply because of the chemical natures of the substance and the fluid (which may also contain modifiers), or because of the particular operating conditions(such as temperature and pressure) being used, problems can arise. Such techniquesare thus restricted in application to substances which are poorly soluble or completelyinsoluble in the chosen supercritical fluid.

A supercritical fluid which is commonly used in particle formation techniquesis supercritical carbon dioxide, which is relatively inexpensive, non-toxic and hasconvenient critical temperature and pressure values. For this particular supercriticalfluid, it is generally non-polar or low polarity substances which cause problems, beingeither very or at least reasonably soluble in it. Thus, for instance, low molecularweight lipophilic materials cannot easily be formed into particles using supercriticalcarbon dioxide.

In the past, such problems have been overcome either by altering the operatingconditions to reduce solubility of the target substance in the supercritical fluid (it isnot always possible, however, to alter the conditions sufficiently to achieve that), orby using a different technique altogether for particle formation. A process known asRESS (Rapid Expansion of Supercritical Solution), for instance, may be used toprecipitate a substance of interest bydissolving it in a supercritical fluid and thenrapidly expanding the resulting solution. However, RESS is generally a less accurateand reliable technique than techniques such as SEDS, allowing less control over thecharacteristics of the particles formed.

Alternatively, one might attempt to use a different supercritical fluid as theanti-solvent, but it can often be very difficult to select a supercritical fluid which isnot only an anti-solvent for the target substance but also capable of dissolving thesolvent vehicle - both requirements need to be met for the fluid to be useable.Supercritical nitrogen, for instance, would act as an anti-solvent for the low molecularweight lipophilic materials which cannot be processed using supercritical carbondioxide, but most conventional organic solvents are insoluble in supercritical nitrogen,so the choice of vehicle would be extremely limited.

It would therefore be advantageous if SEDS, and other similar supercriticalfluid particle formation processes, could be modified to be used under conditionswhere the target substance is soluble in the chosen supercritical fluid. In particular, itwould be desirable to be able to use supercritical carbon dioxide in the production oflipophilic and other low polarity materials. The present invention thus aims tofacilitate the use of supercritical anti-solvents for an even greater number of targetsubstances, and hence to overcome a technical problem in, and widen the field ofapplication for, an already very useful technology.

Statements of the invention

According to a first aspect of the present invention, there is provided a method for forming particles of a target substance, the method involving:

(a) preparing a target solution, containing the target substance dissolved in avehicle which is either a near-critical fluid or a first supercritical fluid;

(b) introducing the target solution into a particle formation vessel; and

(c) contacting the target solution, in the particle formation vessel, with a secondsupercritical fluid under conditions which allow the second supercritical fluid to causeprecipitation of particles of the target substance from the target solution;

wherein the second supercritical fluid is miscible or substantially misciblewith the vehicle and is a fluid in which the target substance is insoluble orsubstantially insoluble.

The vehicle should be soluble or at least partially, preferably substantially,soluble in the second supercritical fluid. The fluids can then dissolve in one anotherby a rapid diffusion process, causing the target substance to "crash out" from itssolution. The second supercritical fluid must not be capable, to anysignificant degree, of dissolving the substance itself as the particles are formed. Inother words, it must be chosen so that the target substance is for all practical purposesinsoluble (preferably having a solubility below 10^-3 mole %, more preferably below10^-5 mole %) in it, under the chosen operating conditions and taking into account anysupercritical fluid modifiers present.

By "miscible" is meant generally that the two fluids are miscible in allproportions under the operating conditions used, and "substantially miscible"encompasses the situation where the two fluids can mix sufficiently well, under thoseoperating conditions; as to achieve the same or a similar effect, ie, dissolution of thefluids in one another and precipitation of the target substance.

Using the method of the invention, particles can be formed even of substanceswhich are soluble in a chosen supercritical fluid, by using that supercritical fluid asthevehicle for the substance, and making use of another supercritical fluid (the secondsupercritical fluid) as the anti-solvent to cause particle precipitation. The targetsubstance is able to dissolve in the vehicle, but to precipitate out of itwhen the vehicle and the second supercritical fluid mix, without product loss into thesecond supercritical fluid.

In previous literature relating to similar particle formation techniques (such asSEDS), it has never been proposed to use as the vehicle anything other than aconventional solvent, in particular not a solvent which is itself in a supercritical ornear-critical condition. However, such vehicles can often have a higher capability fordissolving in the chosen second supercritical fluid than could conventional organicsolvents. Accordingly, they enable the use of processes such as SEDS with new types of target substance.

The method of the invention can be used in any situation where the targetsubstance would be too soluble in the chosen (pure or modified) supercritical fluid toinitiate particle formation, whether because of the chemical natures of the substanceand the fluid or because of the operating conditions used (pressure and temperature,for instance, can significantly affect the solubility of a substance in a supercriticalfluid).

The method is in many ways analogous to the SEDS process of particleformation, in that the anti-solvent (the second supercritical fluid) and vehicle mix withand dissolve in one another, causing removal of the vehicle from the target substance.Accordingly, much of the technical information contained in WO-95/01221 and WO-96/00610,as to the execution of SEDS, can also be applicable when carrying out thepresent invention.

In the following description, the term "supercritical fluid" means a fluid at orabove its critical pressure (P_c) and critical temperature (T_c) simultaneously. Inpractice, the pressure of the fluid is likely to be in the range between 1.01 and 7.0 ofits critical pressure, and its temperature in the range between 1.01 and 4.0 of itscritical temperature (in Kelvin). However, some fluids (eg, helium and neon) haveparticularly low critical pressures and temperatures, and may need to be used underoperating conditions well in excess of those critical values, such as up to 200 timesthe relevant critical value.

Generally, higher temperatures (eg, between about 2 and 4 times the relevantcritical temperature) are preferred, particularly for the second supercritical fluid, sincethey lower the fluid viscosities and can hence improve their dispersing effect on, ordispersion by, the other fluids present In practice the only real constraint on the fluidtemperatures is that they should be below the melting point(s) of the solidsubstance(s) present, in particular of the target substance.

The term "near-critical fluid" encompasses both high pressure liquids, whichare fluids at or above their critical pressure but below (although preferably close to)their critical temperature, and dense vapours, which are fluids at or above their criticaltemperature but below (although preferably close to) their critical pressure.

By way of example, a high pressure liquid might have a pressure betweenabout 1.01 and 7 times its P_c, and a temperature between about 0.5 and 0.99 times itsT_c. A dense vapour might, correspondingly, have a pressure between about 0.5 and0.99 times its P_c, and a temperature between about 1.01 and 4 times its T_c.

The term "vehicle" means a fluid which is able to carry a solid or solids insolution. A vehicle may be composed of one or more componentfluids. In the present invention, the vehicle and the second supercritical fluid shouldbe chosen so that the latter can act, under the operating conditions used, to reduce the capability of the vehicle to carry the target substance, when the two fluids come intocontact

The term "supercritical solution" means a solution of a substance and/oranother fluid, in a supercritical fluid, the solution itself being in a supercritical state.The term "near-critical solution" means a solution, itself in a near-critical state, of asubstance and/or fluid in a near-critical fluid.

The terms "disperse" and "dispersion" refer generally to the transfer of kineticenergy from one fluid to another. They usually imply the formation of droplets, or ofother analogous fluid elements, of the fluid to which the kinetic energy is transferred,typically of the solution or suspension of the target substance and/or of the vehicle.

The target substance may be any substance which needs to be produced inparticulate form. It may be a substance for use in or as a pharmaceutical. However, itmay also be a material of use in the ceramics, explosives or photographic industries; afoodstuff; a dye; a coating; etc. It may be organic or inorganic, monomeric orpolymeric. In each case, the principle behind the method of the invention remains thesame; the technician need only adjust operating conditions and choose appropriatefluids in order to effect proper control over the particles being formed.

The target substance may be in a single or multi-component form. Theparticulate product formed from the substance may also be in a multi-component form- it could for instance comprise an intimate mixture of two materials, or one materialin a matrix of another, or one material coated onto a substrate of another, or othersimilar mixtures. Such products may be made from solutions or suspensionscontaining only single component starting materials (more than onesolution/suspension, not all of which need contain a supercritical or near-critical fluidvehicle, may be introduced into the particle formation vessel with the secondsupercriticai fluid). The particulate product may also be a substance formed from anin situ reaction (ie, immediately prior to, or on, the target solution contacting thesecond supercritical fluid) between two or more reactant substances, each carried byan appropriate vehicle. To produce such multi-component products, the secondsupercritical fluid may itself be used to carry a component such as a reactant or acarrier material, intended to be incorporated into the final product.

Modifications involving the use of in situ reactions and/or more than onesolution or suspension of a target substance, are described in connection with SEDS inWO-95101221 and WO-96/00610, and can also be applied when carrying out thepresent invention.

The target substance will typically be non-polar or of fairly low polarity, inwhich case the vehicle should also be of low polarity and the second supercriticalfluid may have a relatively high polarity. However, the reverse may also apply, ie,polar target substance with polar vehicle and relatively low polarity second supercritical fluid.

The vehicle is preferably, although not necessarily, a supercritical fluid. Itmay for instance be carbon dioxide, nitrogen, nitrous oxide, sulphur hexafluoride,xenon, ethylene, chlorotrifluoromethane, ethane or trifluoromethane, in an appropriatecondition with respect to its P_c and T_c- A particularly preferred vehicle is carbondioxide (P_c = 74 bar; T_c = 31°C), preferably supercritical carbon dioxide. The choicenaturally depends on the target substance; in the case of low polarity or non-polarsubstances, in particular low molecular weight lipophilic substances, supercriticalcarbon dioxide is again an appropriate vehicle.

The vehicle may include more than one supercritical or near-critical fluid,and/or other fluids such, as conventional solvents, provided that it has, overall, thenecessary solubility properties vis-a-vis the target substance and the secondsupereritical fluid. It may optionally contain one or more modifiers or co-solvents. Amodifier (or co-solvent) may be described as a chemical which, when added to a fluidsuch as a supercritical or near-critical fluid, changes the intrinsic properties of thatfluid at or around its critical point, in particular its ability to dissolve other materials.Suitable modifiers include water, and conventional organic solvents such as methanol,ethanol, isopropanol or acetone. When used, a modifier preferably constitutes notmore than 40 mole %, more preferably not more than 20 mole %, and most preferablybetween 1 and 10 mole %, of the vehicle.

The choice of vehicle in any particular case will depend on the nature of thetarget substance, on the second supercritical fluid and on other practical criteriaincluding those governing the desired end product. The target substance issoluble in the vehicle, preferably having a solubility in the vehicle of 10^-4 mole % orgreater, under the chosen operating conditions (pressure, temperature and modifierspresent).

The second supercritical fluid may be for instance supercritical carbondioxide, nitrogen, nitrous oxide, helium, sulphur hexafluoride, xenon, ethylene,chlorotrifluoromethane, ethane, trifluoromethane or a mixture of any of these. Aconvenient second supercritical fluid is supercritical nitrogen (P_c = 33.9 bar, T_c =minus 147°C), which is an anti-solvent for many solids, including many non-polar orlow polarity substances. Supercritical noble gases, for example helium, can also beeffective anti-solvents for many target substances.

The use of supercritical or near-critical carbon dioxide as the vehicle andnitrogen, as the second supercritical fluid is particularly effective, since supercriticalnitrogen is miscible with, and can readily dissolve, supercritical carbon dioxide.

The second supercritical fluid may also contain one or more modifiers, of thesame general type and in the same proportions as can the vehicle. As mentionedabove, it may contain further target substances or reactants for introduction into the particle formation vessel.

The choice of suitable operating conditions to allow particle precipitation tooccur will be well within the capabilities of the person skilled in this art. Generally,the conditions in the particle formation vessel must be such that the secondsupercritical fluid remains in the supercritical state whilst in the vessel. The fluidmixture formed when the target solution comes into contact with the secondsupercritical fluid should also, for at least one of its constituent fluids (usually thesecond supercritical fluid, which in general will be the major constituent of themixture), be in a supercritical state at the time of particle formation. The mixtureshould at that time comprise asingle-phase mixture of the vehicle and the secondsupercritical fluid, otherwise the particulate product might be distributed between twoor more fluid phases, in some of which it might be able to redissolve. This is why thesecond supercritical fluid needs to be miscible or substantially miscible with thevehicle.

The pressure and temperature inside the particle formation vessel may be thesame as or different to those of the target solution about to be introduced intoit

Under these conditions, particle precipitation will generally be immediate, oreffectively immediate, when the target solution and the second supercritical fluidcome into contact. This allows the rapid formation of pure, dry particulate products.The exact pressures and temperatures needed to achieve this situation depend ofcourse on the nature of the second supercritical fluid and on the target substance, thevehicle and any other fluids being used.

The method by which the target solution and the second supercriticalfluid are introduced'into the particle formation vessel may be analogous to that usedin SEDS, or in a modified version thereof such as is described in WO-95/01221 orWO-96/00610.

Thus, the target solution and the second supercritical fluid may be co-introducedinto the particle formation vessel through a fluid inlet means which allowsthe second supercritical fluid (physically) to mix and disperse the target solution,
at the same time as it (chemically) causes precipitation from it. This canprovide a very high degree of control over the particles formed.

To achieve this, the fluid inlet means may allow both the target solutionand the second supercritical fluid to enter the particle formation vessel at the same orsubstantially the same point, which is also the same as, or substantially the same as,the point where they meet. Preferably, the inlet means is so arranged that the secondsupercritical fluid can act, at that same point or substantially that same point, tofacilitate intimate mixing of the fluids, and preferably to break up thetarget solution into individual fluid elements (the precursors to the eventual particles), by the transfer of kinetic energy from the second supercritical fluid to thetarget solution. Preferably, particle precipitation is allowed to occur at orsubstantially at the point where the two fluids enter the particle formation vessel.

The fluid inlet means may be of the type which allows "pre-filming" or"sheathing" of at least one of the fluids to occur, immediately prior to its dispersion byan impinging flow of another fluid introduced through the inlet means. For instance,the inlet means can be used to cause pre-filming of the target solutionimmediately prior to its dispersion by the second supercritical fluid. This means thatthe dimensions of the inlet passages of the inlet means, and the relative positions oftheir outlets, must be such that a fluid entering through one passage is formed, as itreaches the outlet of that passage, into a thin film or sheath, by its contact with, say,the lip of an adjacent passage outlet. This film or sheath can then be stretched, andultimately dispersed into separate fluid elements, when it comes into contact with anoncoming stream of a fluid in another inlet passage. Clearly, the thickness of the filmor sheath, and hence the sizes of the fluid elements formed on dispersion, will dependto a large extent on the relative flow rates of the fluids, and also on the inlet passagedimensions.

In one embodiment of the invention, the second supercritical fluid and thetarget solution are co-introduced into the particle formation vessel withconcurrent directions of flow, preferably with coaxial or substantially coaxial flows,such as using a multi-passage coaxial nozzle. Such a nozzle has an outlet endcommunicating with the interior of the particle formation vessel, and two or morecoaxial, conveniently concentric, passages which terminate adjacent or substantiallyadjacent one another at the outlet end, at least one of the passages serving to introducea flow of the second supercritical fluid into the particle formation vessel, and at leastone of the passages serving to introduce a flow of the target solution.

Aspects of such a coaxial nozzle may be as described in WO-95/01221 andWO-96/00610. For instance, the opening at the outlet end (tip) of the nozzle willpreferably have a diameter in the range of 0.005 to 5 mm, more preferably 0.05 to 2mm, most preferably between 0.1 and 0.5 mm, for instance about 0.1, 0.2, 0.3 or 0.35mm. The angle of taper of the outlet end (with respect to the longitudinal axis of thenozzle) will depend on the desired velocity of the fluids introduced through thenozzle; a change in the angle may be used. for instance, to increase the velocity of thesecond supercritical fluid and hence to increase the amount of its physical contactwith the target solution, leading to more efficient fluid mixing. Typically, theangle oftaper will be in therange 10° to 60°, preferably between 10° and 50°, morepreferably between 20° and 40°, and most preferably about 30°. Alternatively, theoutlet need not be tapered at all.

The nozzle may be made of any appropriate material, for example stainless steel. It may have any appropriate number of coaxial passages (preferably two orthree), some of which may be used to introduce additional reagents into the particleformation vessel. One or more of the passages may be used to introduce two or morefluids at the same time, and the inlets to such passages may be modified accordingly.

The internal diameters of the coaxial passages may be chosen as appropriatefor any particular case. Typically, for a three-passage nozzle, the ratio of the internaldiameters of the outer and the inner passages may be in the range from 2 to 10,preferably between 2 and 5, more preferably between 3 and 4. The ratio of theinternal diameters of the outer and intermediate passages may be in the range from1.01 to 5, preferably between 1.2 and 3. For a two-passage nozzle, the ratio of theinternal diameters of the outer and inner passages may be in the range from 1 to 10,preferably between 2 and 6, more preferably between 2 and 4.

The outlets of two or more of the passages may be relatively staggered alongthe longitudinal axis of the nozzle, ie, one passage may terminate slightly (preferablybetween about 0.05 and 10 mm, more preferably between about 0.05 and 1 mm)upstream or downstream, in use, of another. For instance, the outlet of an innerpassage may terminate slightly upstream of that of a surrounding passage, to allow adegree of internal mixing between fluids introduced through the two passages.

The use of such coaxial nozzles as the fluid inlet means can minimise contactbetween the formed particles and the vehicle in the region of the inlet means, whichcontact could reduce control of the final product size, shape and yield. Extra controlover the size of the dispersed vehicle fluid elements, in addition to that provided bythe design of the inlet means, may be achieved by controlling the flow rates of thesecond supercritical fluid and the target solution into the particle formationvessel.

In the method of the invention, however, co-introduction of the fluids into theparticle formation vessel is not essential. The target solution and the secondsupercritical fluid may be introduced through separate inlet means and/or at differentlocations in the vessel, and/or at different angles to one another. They may even beintroduced into the vessel at different times. All that is needed is that they come intocontact with each other under conditions which allow the second supercritical fluid tocause particles to precipitate out of the target solution. Ideally, however, theyare introduced in a way which ensures efficient fluid mixing immediately prior to, andat, the point of particle formation, and the fluid inlet means is/are preferably designedto assist in this.

In the method of the invention, the relative flow rates of the fluids introducedinto the particle formation vessel may be used to influence the characteristics, inparticular the size and/or size distribution, of the particles formed. This is particularlyrelevant when using the action of the second supercritical fluid to disperse the target solution. Preferably, the flow rate of the second supercritical fluid,measured at or immediately prior to its contact with the target solution, ishigher than that of the target solution -this can lead to the formation ofgenerally smaller fluid elements (eg, droplets) of the target solution, andhence relatively small particles, having a narrow size distribution, when the secondsupercritical fluid causes precipitation out of the fluid elements.

In some cases, the flow rate of the second supercritical fluid might be up to100 times greater than that of the target solution; preferably between 5 and 30times greater, more preferably between 10 and 20 times greater. (These figures againrefer to fluid flow rates measured at or immediately prior to the fluids coming into.contact with one another.) It will in general be chosen to ensure an excess of thesecond supercritical fluid over the vehicle when the fluids come into contact, tominimise the risk of the vehicle re-dissolving and/or agglomerating the particlesformed. At that point, the amount of the second supercritical fluid relative to that ofthe vehicle must naturally be sufficient to cause particles of the target substance toprecipitate from the target solution. Typically, the vehicle may constitutebetween 1 and 80 mole %, preferably between 1 and 20 mole %, more preferablybetween 1 and 5 mole %, of the fluid mixture formed.

The method thus preferably involves controlling the flow rate of one or moreof the fluids to influence the characteristics of the particulate product. The flow rateof the target solution, as well as that of the second supercritical fluid, may affectthese characteristics.

The flow rates of the second supercritical fluid and the target solution,together with the concentration of the target substance in its solution,are ideally selected so as to minimise the risk of particle precipitation in, and henceblockage of, the fluid inlet means or other upstream apparatus parts. For the samereason, back-flow of the second supercritical fluid, into the supply line for thetarget solution, is also ideally reduced or eliminated This may be done forinstance by installing a one-way valve in the target solution supply line,upstream of and preferably immediately before the point of contact between thesecond supercritical fluid and the target solution. Instead or in addition, thesupplied pressure of the target solution, to the particle formation vessel, maybe maintained in excess of that of the second supercritical fluid. It is also preferableto maintain the temperature of the target solution, prior to its introduction intothe particle formation vessel, in excess of that of the second supercritical fluid.

The fluids are ideally introduced into the particle formation vessel with asmooth, continuous and preferably pulse-less or substantially pulse-less flow. Thisagain helps prevent draw-back of fluids. Conventional apparatus may be used toensure such fluid flows.

The densities of the fluids used will depend on the operating conditions, whichin turn will be selected according to the natures of the species present and of thedesired end product In the case where supercritical carbon dioxide is used as thevehicle, its density might typically be between 0.1 and 0.9 g/ml. Supercriticalnitrogen might typically be used as the second supercritical fluid at a density ofbetween 0.01 and 0.05 g/ml, preferably around 0.02 g/ml.

Depending on the nature of the target substance, two main embodiments of themethod of the invention are likely to be preferred:

a) where the target substance is highly soluble, preferably freely solublein the vehicle, it is dissolved directly in the vehicle and the resultantsolution is then introduced into the particle formation vessel to contactthe second supercritical fluid. In order to dissolve the substance in thevehicle, it is preferably charged into a vessel through which a flow ofthe vehicle is passed to form a saturated solution.

b) where the target substance is less than freely soluble in a chosenvehicle, a solution of the substance may firstly be made up usinganother ("primary") solvent in which it is more soluble. This solutionis then itself dissolved in the chosen vehicle, forinstance by dispersing the solution into a fluid mixing vessel togetherwith the vehicle (this dispersion process may also be carried out usingan inlet means of the type suitable for use in the particle formation stepof SEDS). The resulting supercritical or near-critical solution is thencontacted with the second supercritical fluid in the particle formationvessel.

When using embodiment (a), the target substance can preferably form a stablesingle-phase solution in the vehicle, at a target substance:vehicle weight ratio of atleast 1:1, under the relevant operating conditions. The embodiment may however beused for less soluble target substances, perhaps even at solubilities as low as 10^-1 mole%, provided sufficient time can be allowed for their dissolution in the vehicle.

For embodiment (b), the solubility of the target substance in the vehicle, underthe relevant operating conditions, may typically be less than about 10^-2 mole %,possibly less than about 10^-4 mole %. Embodiment (b) may also of course be used fortarget substances whichare freely soluble in the chosen vehicle.

In embodiment (b), the "vehicle" is effectively, at the time of particleformation, the initially chosen fluid plus the primary solvent (which can also be seenas the chosen fluid plus a modifier). The embodiment could be of use where the ,target substance, although only slightly soluble in a chosen fluid, is much moresoluble in that fluid when a few volume percent of a modifier has been added to itThe "vehicle" need only be in a supercritical or near-critical state at the point where it contacts the second supercritical fluid, and not necessarily at the point where it ismixed with the primary solvent.

The present invention also provides, according to a second aspect, apparatusfor use in carrying out the above described embodiment (b). This apparatus includes aparticle formation vessel; means for controlling the temperature and pressure in theparticle formation vessel at desired levels; a fluid mixing vessel; means for controllingthe temperature and pressure in the fluid mixing vessel at desired levels; first fluidinlet means for introducing into the fluid mixing vessel a vehicle and a solution of atarget substance in a primary solvent, so as to form in the fluid mixing vessel asolution of the target substance and the primary solvent in the vehicle; and secondfluid inlet means for introducing the solution thus formed, preferably together with asecond supercritical fluid, into the particle formation vessel.

Each of the first and second fluid inlet means may be of the type described inconnection with the first aspect of the invention.

The invention will now be described by way of example only, with referenceto the accompanying illustrative drawings, of which:-

Figure 1 illustrates schematically how a method in accordance with the firstaspect of the invention may be carried out;

Figure 2 illustrates schematically how an alternative method may be carriedout in accordance with the invention;

Figures 3-6 are particle size distribution curves for products obtained inaccordance with the invention (see experimental Examples 1a, 1b, 1c and 2);

Figures 7 and 8 are DSC (differential scanning calorimetry) profiles for thestarting material and the product respectively of Example 1a;

Figures 9-13 are XRPD (X-ray powder diffiaction) patterns for, respectively,the starting material and the product of Example 1a, the products of Examples 3a and3b and the product of Example 4;

Figures 14-18 are SEM (scanning electron microscope) photographs ofnicotinic acid produced, respectively, in Example 5a, using SEDS, by a conventionalcrystallisation and micronisation process and in Examples 5b and 5c; and

Figures 19-21 are XRPD patterns for, respectively, nicotinic acid produced bycrystallisation and micronisation and the products of Examples 5a and 5b.

Detailed description

Figure 1 shows, schematically only, how a method according to the inventionmay be carried out. In the example described, the vehicle used is supercritical. carbondioxide, and the second supercritical fluid is supercritical nitrogen. The target substance is soluble in supercritical carbon dioxide.

Carbon dioxide from source 1 is passed throughcooler 2, pump 3 andheatexchanger 4, to take it into its supercritical state. It then passes throughsample vessel5, containing a target substance from which particles are to be formed. The targetsubstance has been charged into thevessel 5 with glass beads, to form a bed exposingas high as possible a surface area to the supercritical carbon dioxide, and also toeliminate the risk of channelling of the carbon dioxide.

The carbon dioxide dissolves the substance and the resulting solution passes,viapressure regulator 6, through a two-component inlet nozzle 7 intoparticleformation vessel 8.

Supercritical nitrogen is also fed into thevessel 8 via nozzle 7, as shown at 9.The pressure and temperature insidevessel 8 are controlled by means of surroundingoven 10 and automated back pressure regulator 11.

At the nozzle outlet, the supercritical nitrogen contacts the solution of thetarget substance in supercritical carbon dioxide, dissolves in the carbon dioxide andcauses precipitation of particles of the target substance, which collect in the particleretaining device (such as a filter or cyclone) 12. Supercritical conditions aremaintained in the vessel, allowing the fluids (ie, a supercritical mixture of carbondioxide and nitrogen) to be removed to vent 13, via the back pressure regulator 11 andaflow meter 14.

The nozzle 7 preferably has two coaxial passages, one for introduction of thecarbon dioxide/target substance solution, and one for introduction of the supercriticalnitrogen. It preferably allows these two fluids to be introduced into the vessel withconcurrent directions of flow, in such a way that they meet and enter the vessel atsubstantially the same point, and preferably also in such a way that the mechanicalenergy of the supercritical nitrogen contributes to the efficient mixing of the twofluids at their point of contact.

In carrying out the invention according to Figure 1, it is desirable to form asaturated solution of the target substance in the supercritical vehicle prior tointroducing the solution into the particle formation vessel. The flow rate of thesecond supercritical fluid should be high with respect to that of the vehicle/targetsubstance solution.

The apparatus shown schematically in Figure 2 is for use in producingparticles of a target substance which is less than freely soluble in the chosen vehiclebut much more soluble in the vehicle once modified with another solvent. Forinstance, a slightly polar target substance might not be highly soluble in supercriticalcarbon dioxide, but much more soluble in a mixture of supercritical carbon dioxideand a few mole percent of a polar modifier.

In the Figure 2 system, the vehicle (in this example carbon dioxide) is fed fromsource 21 through cooler 22, pump 23 andheat exchanger 24, to bring it into asupercritical state. It is then fed into a mixingvessel 25, together with a solution ofthe target substance in an appropriate primary solvent (fromsource 26, via pump 27).The solution and the supercritical carbon dioxide enter the mixing vessel through atwo-passagecoaxial nozzle 28, of the type described above, which allows the kineticenergy of the carbon dioxide to act to disperse the solution, ensuring thorough mixingof the two fluids when they meet at the nozzle outlet. This yields a solution of thetarget substance in the carbon dioxide and the primary solvent, ie, effectively asolution of the target substance inmodified supercritical carbon dioxide. The"modified" solution is maintained in a supercritical state by controlling the pressureand temperature inside the mixing vessel, usingpressure regulator 29 and surroundingoven 30.

The supercritical solution then passes to asecond nozzle 31, through which itis co-introduced into aparticle formation vessel 32 with a second supercritical fluidflowing with a relatively high flow rate.Nozzle 31 is of the same general form asnozzle 28. At the same time, or substantially the same time, as the fluids meet andenter the vessel, the second supercritical fluid dissolves in the carbon dioxide vehicle,reduces its capacity for the target substance and thus causes precipitation of particleswhich can be collected in theparticle retaining device 33. Again, supercriticalconditions are maintained in thevessel 32 using backpressure regulator 34 and theoven 30.

The fluid mixture'which remains after particle precipitation can be vented atthe bottom of theparticle formation vessel 32, viaflow meter 35.

If difficulties arise in pumping low boiling point liquefied gases such asliquefied nitrogen, it may be convenient, at least for the laboratory scale, to use gaspumps or cylinders providing an appropriate level of pressure. The gases can simplybe vented from their higher pressure cylinders into the particle formation vessel whichis at a lower pressure. Their flow can be controlled by a needle valve. Cryogenicpumps may also be used if the circumstances warrant.

It is important to keep the rate of fluid addition as constant as possible and toachieve efficient fluid mixing throughout the particle formation process. To acquiregood control over the rate of addition of, for instance, supercritical nitrogen underhigh pressures, gas cylinders or boosters may he used. Of these, motor driven gasboosters tend to be more effective than high pressure gas cylinders, since they canoffer:

higher operation pressures and flow outputs;
better control over flow rates high pressure cylinders suffer from continuouspressure drop during operation and a bank at cylinders is often needed to compensatefor this pressure reduction during a long particle formation procedure): and
longer fluid delivery times.

Laboratory scale gas boosters are available from, for instance, Stansted PowerFluid Ltd (Essex, UK) and can deliver up to 40 standard litres per minute at pressuresexceeding 350 bar.

Experimental examples

The following examples demonstrate how the method of the present inventionmay be used to produce a range of target materials, some of which would beincompatible with a conventional SEDS process, whilst allowing a high degree ofcontrol over the product properties.

Example 1

This experiment made use of the system described in connection with Figure1. The drug ibuprofen was the target substance, chosen for its high solubility in bothpure and modified supercritical carbon dioxide. The vehicle was supercritical carbondioxide and the second supercrilical fluid was supercritical nitrogen. The method ofthe invention was used to produce particles of the drug (Example 1a) and to controlthe size of those particles by varying the flow rates of the fluids involved (Examples1b and 1c).

Example 1a

I g of ibuprofen was mixed with glass beads (200-301) µm acid washed(Sigma, UK)) and introduced into a 10 ml Keystone vessel (the sample vessel 5) toform a uniform bed. The bed was sandwiched between two filters (average pore size2 µm) to eliminate the risk of physical entrainment of drug particles in the carbondioxide flow. The sample vessel was provided with a pressure regulator independentof that of theparticle formation vessel 8.

The fluids were introduced into the particle formation vessel using a two-passagecoaxial nozzle of the preferred type described above, having a 0.1 mmdiameter outlet. The nozzle ensured thorough mixing of the fluids at their point ofcontact, ie, at their point of entry into the vessel. The conditions in the vessel weresuch that particle formation occurred simultaneously, or substantially so, on the fluidsmeeting and entering the vessel.

1 ml/min of carbon dioxide (measured at the pump head) was pumped into thesample vessel containing the ibuprofen bed, which was maintained at 130 bar. Theresultant supercritical solution, at the same flow rate, was introduced into the particleformation vessel via the outer passage of the nozzle, and supercritical nitrogen wasintroduced through the inner passage. The nitrogen flow rate, measured at the flow meter 14 (ie, after the back pressure regulator) at ambient conditions, was keptconstant through all experiments at 10 l/min.

The pressure inside the particle formation vessel (a 50 ml Keystone vessel)was set at 60 bar. The oven temperature was 40°C.

At the end of the experiment a fine, fluffy white powder was collected in theretaining device 12 and stored free from moisture for subsequent analysis.

Particle size analysis of the product was carried out using theAerosizer/Aerodisperser system (API, USA). The results, in the form of a particlesize distribution curve, are shown in Figure 3 and summarised in Table 1 below. Themean particle diameter, by volume, was about 21 µm.

% UNDER	SIZE	% UNDER	SIZE
5%	9.241	55%	24.87
10%	11.68	60%	26.16
15%	13.57	65%	27.48
20%	15.21	70%	28.83
25%	16.74	75%	30.25
30%	18.21	80%	31.76
35%	19.62	85%	33.44
40%	20.99	90%	35.44
45%	22.30	95%	38.65
50%	23.58
Mean size:	21.68
Standard deviation:	1.569

Example 1b

Example 1a was repealed, but with the carbon dioxide flow rate increasedfrom 1 to 4 ml/min (at the pump head). All other operating conditions remained thesame,

The product was a fine fluffy white powder. Analysis using theAerosizer/Aerodisperser system yielded the particle size distribution curve shown inFigure 4 and summarised in Table 2. The mean particle diameter, by volume, was about 14 µm.

% UNDER	SIZE	% UNDER	SIZE
5%	5.920	55%	16.18
10%	7.411	60%	17.07
15%	8.631	65%	17.99
20%	9.720	70%	18.94
25%	10.75	75%	19.95
30%	11.72	80%	21.05
35%	12.66	85%	22.25
40%	13.56	90%	23.67
45%	14.43	95%	25.84
50%	15.30
Mean size:	14.12
Standard deviation:	1.586

Example 1c

Again Example 1a was repeated, but this time with a carbon dioxide flow rateof 8 ml/min (at the pump head). The product was again a fine fluffy white powder,which when analysed (see Figure 5 and Table 3) showed a mean particle diameter, byvolume, of about 8 µm.

% UNDER	SIZE	% UNDER	SIZE
5%	3.762	55%	9.180
10%	4.462	60%	9.849
15%	5.018	65%	10.55
20%	5.518	70%	11.28
25%	5.999	75%	12.06
30%	6.479	80%	12.92
35%	6.966	85%	13.88
40%	7.471	90%	14.96
45%	8.004	95%	16.35
50%	8.572
Mean size:	8.333
Standard deviation:	1.589

Example 1 - Conclusions

These experiments show that varying the flow rate of the vehicle, in which thetarget substance is dissolved, can be used to influence the size of particulate productsmade according to the invention. Here, increasing the flow rate led to reduced particlesizes.

Example 2

This experiment made use of the system described in connection with Figure2. Ibuprofen was dissolved in a conventional solvent (methanol) prior to introductionof the supercritical carbon dioxide vehicle. Supercritical nitrogen was used to causeparticle precipitation.

A 10% w/v solution of ibuprofen in methanol was pumped at 0.05 ml/min intothe mixingvessel 25 via the two-passagecoaxial nozzle 28, together with supercriticalcarbon dioxide at a flow rate of 4 ml/min (measured at the pump head). The nozzleoutlet had a diameter of 0.2 mm. The mixing vessel (24 ml, Keystone) was kept at150 bar and 40°C.

The resultant solution, mixed and dispersed by the action of the supercriticalcarbon dioxide, was then fed to the particle formation vessel 32 (50 ml, Keystone), viathe inner passage of another two-component nozzle 31, this time with a 0.1 mmdiameter outlet. Supercritical nitrogen was introduced through the outer passage ofthe same nozzle, with a flow rate (measured at ambient temperature) of 10 l/min.

The product was a fine fluffy white powder, collected in thevessel 32.Analysis using the Aerosizer revealed (see Figure 6 and Table 4) a mean particlediameter, by volume, of about 14 µm.

% UNDER	SIZE	% UNDER	SIZE
5%	5.815	55%	16.10
10%	7.327	60%	17.06
15%	8.537	65%	18.06
20%	9.612	70%	19.10
25%	10.62	75%	20.20
30%	11.57	80%	21.41
35%	12.49	85%	22.79
40%	13.38	90%	24.42
45%	14.26	95%	26.78
50%	15.17
Mean size:	14.13
Standard deviation:	1.611

Polymorphic form of the products

The products of Examples 1 and 2, when examined using DSC (differentialscanning calorimetry) and XRPD (X-ray powder diffraction) techniques, were allfound to be of the same polymorphic form as the starting material. By way ofdemonstration, Figure 7 is a DSC profile for the ibuprofen used as the startingmaterial; Figure 8 is a DSC profile for the product of Example 1a. Figures 9 and 10are XRPD spectra for the starting material and the product of Example 1arespectively.

These data show that the methods of the present invention may be used toform particles of target substances without compromising their purity or theircrystalline form.

Example 3

In this experiment, the system described in connection with Figure 2 was usedto prepare particulate salicylic acid, which is soluble in supercritical carbon dioxideprovided a small amount of a polar modifier is also present. Particle formation wascarried out using two different salicylic acid solutions as the starting materials (Examples 3a and 3b).

Example 3a

A 3% w/v solution of salicylic acid in methanol was introduced at a flow rateof 0.2 ml/min into the mixing vessel 25 (in this case, a 5 ml Keystone vessel), via theinner passage (internal diameter 0.15 mm, external diameter 0.30 mm) of a two-passagecoaxial nozzle. Supercritical carbon dioxide was introduced through theouter passage (internal diameter 0.35 mm) at a flow rate of 9 ml/min measured at thepump head. The nozzle outlet diameter was 0.35 mm, and the outlet of the innerpassage terminated 0.2 mm upstream of that of the outer passage. The mixing vesselwas maintained at 200 bar and 50°C.

Since the solubility of the acid in the supercrilical fluid increases dramaticallyin the presence of a few percent of the polar modifier methanol, little or no particleformation was expected to occur in the mixing vessel.

The supercritical solution thus formed was introduced into the particleformation vessel 32 (a 50 ml Keystone vessel) via the inner passage of another two-passagecoaxial nozzle of the same dimensions as that used in the mixing vessel,together with supercritical nitrogen flowing at 10 l/min (measured at atmosphericconditions) through the outer passage. The particle formation vessel was alsomaintained at 200 bar and 50°C.

At the end of the run a crystalline white powder was collected in thevessel 32.Its XRPD pattern is shown in Figure 11.

Example 3b

Example 3a was repeated but starting with a 2% w/v solution of salicylic acidin dichloromethane. Again the dichloromethane acts as a polar modifier, increasingthe solubility of the acid in the supercritical carbon dioxide. The fluid flow rates werethe same as used in Example 3a, but the operating conditions inside the mixingvessel25 and theparticle formation vessel 32 were 200 bar and 65 °C.

The product was again a fine fluffy white powder, crystalline in form (seeFigure 12).

Example 4

The system described in connection with Figure 1 was used to produce thedrug ketoprofen in particulate form.

0.5 g of ketoprofen was mixed with acid washed glass beads, average diameter200-300 µm and packed into a bed inside the sample vessel 5 (in this case, a 5 mlKeystone pressure vessel fitted with 0.5 µm sinters). Supercrilical carbon dioxidewas then introduced through a frit/sinter at the bottom of the vessel, at a flow rate of 9 ml/min measured at the pump head. The vessel was maintained at 200 bar and 50°CThe sizes of the glass beads were selected to enhance the drug surface area availablefor contact with the carbon dioxide and also to discourage caking of the bed.

The supercritical solution (of ketoprofen in carbon dioxide) emerging from thetop of the sample vessel was introduced into the particle formation vessel 8 (50 mlKeystone) through the inner passage of a two-passage coaxial nozzle of the type usedin Example 3, still at a flow rate of 9 ml/min. Supercritical nitrogen was introducedthrough the outer passage at a flow rate of 10 standard litres/min. The pressure andtemperature inside the particle formation vessel were maintained at 200 bar and 50°C.

A fine fluffy white powder was collected in the particle formation vessel. ItsXRPD pattern (Figure 13) confirmed its crystallinity.

Example 5

This experiment demonstrates the successful use of the present invention atrelatively high operating pressures and temperatures. Experiments were carried outunder three different sets of operating conditions (Examples 5a - 5c).

For some target substances, high temperatures and pressures are needed inorder to produce particles having the desired physicochemical properties. By way ofexample, to produce the drug salmeterol xinafoate in the form of its polymorph II,pressures greater than 250 bar and temperatures greater than 85 °C are needed.However, applying such "harsh" working conditions is not appropriate for all targetsubstance/fluid combinations. For instance, nicotinic acid has a relatively lowsolubility in pure and modified supercritical carbon dioxide at pressures below 120bar and temperatures below 90°C; under such conditions supercritical carbon dioxidecould be used as an anti-solvent to precipitate the acid from solution. However, above150 bar the solubility of nicotinic acid in supercritical carbon dioxide increasesdramatically, and an alternative anti-solvent must be found.

In such a case, the method of the present invention may be used to produceparticles of the target substance under the desired high temperature and pressureconditions, despite its solubility, under those conditions, in the first choice ofsupercritical anti-solvent.

Example 5a

A 0.8% w/v solution of nicotinic acid in methanol was introduced at a rate of0.2 ml/min, through the inner passage of a nozzle of the type used in Example 3, intothe mixingvessel 25 of the Figure 2 system. The vessel (5 ml Keystone) wasmaintained at 200 bar and 65°C. Supercritical carbon dioxide was introduced throughthe outer nozzle passage at a flow rate of 9 ml/min measured at the pump head. Theresultant supercritical solution was introduced into a 50 ml Keystone vessel (the particle formation vessel 32) also kept at 200 bar and 65°C, together withsupercritical nitrogen flowing at 10 standard litres/min - the same type of nozzle wasused, the nicotinic acid solution flowing through the inner passage and the nitrogenthrough the outer.

At the end of the run, a fine fluffy white powder was collected in thevessel32. An SEM micrograph of the product (Figure 14) shows it to have a similar particlesize and morphology to that of nicotinic acid prepared using a SEDS process asdescribed in WO-95/01221 (Figure 15), but very different to that of theconventionally crystallised and micronised material (Figure 16). (For the SEDS"control", a 0.8% w/v solution of the acid in absolute ethanol was co-introduced into aparticle formation vessel kept at only 90 bar and 85 °C, via a two-passage coaxialnozzle, with supercritical carbon dioxide as the anti-solvent; the fluid flow rates were0.2 ml/min for the acid solution and 9 ml/min (measured at the pump head) for theanti-solvent. It is of note that the same process carried out at 200 bar and 85 °Cyielded no product at all, all the nicotinic acid being extracted by the supercriticalcarbon dioxide and precipitated at the vent line downstream of the particle formationvessel.)

Example 5b

Example 5a was repeated, but at a higher operating temperature of 85°C. Theproduct was again a fine, fluffy white powder, containing well-faceted microcrystals(as seen in Figure 17) and having a comparable particle size and morphology to thatof the product of Example 5a.

Example 5c

Example 5a was repeated using an operating temperature of 100°C. Theproduct was again a fine fluffy white powder. SEM examination (Figure 18) revealeda similar morphology to that of the Example 5a product, but surprisingly a smallerparticle size. This could be because at higher temperatures the viscosity of thesupercritical nicotinic acid/carbon dioxide solution is lowered and its linear velocity atthe nozzle outlet therefore raised, thus improving its dispersion by the supercriticalnitrogen.

The products of Examples 5a-5c exhibited a high degree of crystallinity, andthe same morphology as a micronised form of nicotinic acid - see the XRPD patternsof Figures 19 (the micronised product), 20 (the product of Example 5a) and 21 (that ofExample 5b).